Search

Artificial neural networks are successfully used to solve a wide variety of scientific and technical problems. The purpose of the study is to increase the efficiency of distributed solutions for problems involving structural-parametric synthesis of neural network models of complex systems based on GRID (geographically disperse computing resources) technology through the integrated application of the apparatus of evolutionary optimization and queuing theory. During the course of the research, the following was obtained: (i) New mathematical models for assessing the performance and reliability of GRID systems; (ii) A new multi-criteria optimization model for designing GRID systems to solve high-resource computing problems; and (iii) A new decision support system for the design of GRID systems using a multi-criteria genetic algorithm. Fonseca and Fleming's genetic algorithm with a dynamic penalty function was used as a method for solving the stated multi-constrained optimization problem. The developed program system was used to solve the problem of choosing an effective structure of a centralized GRID system that was configured to solve the problem of structural-parametric synthesis of neural network models. To test the proposed approach, a Pareto-optimal configuration of the GRID system was built with the following characteristics: average performance–103.483 GFLOPS, cost–500 rubles per day, availability rate–99.92%, and minimum performance–51 GFLOPS. [ABSTRACT FROM AUTHOR]

Modern ships are complex energy systems containing a large number of different elements. Each of these elements is simulated separately. Since all these models form a single system (ship), they are interdependent. The operating modes of some systems influence others, but at the same time, the work of all the systems should be aimed at fulfilling the basic functions of the ship. The work proposes a methodological approach to combining various systems of ships into a single complex model. This model allows combining models of ship systems of various levels (microlevel, macrolevel, metalevel, megalevel). The work provides examples of models of such multi-level energy systems. These are energy systems composed of an electric generator, a diesel engine, a propeller shaft, and algorithms used for operating the common parts of the ship's electric power system and a piston wear process. Analytical, structural, numerical, and object-oriented models were made for these objects. Each of these particular models describes a limited class of problems, has characteristic properties, and a mathematical structure. The work shows how particular models can be interconnected using a set-theoretic description. Particular models are combined into macrolevel models, whose output parameters are quantities that are by no means related. The macrolevel models are interrelated using control models. Control models belong to the metalevel and allow for assigning settings and response thresholds to algorithms used in automation systems. Such a model (megalevel model) allows, ultimately, investigating the dynamics of the entire system as a whole and managing it. [ABSTRACT FROM AUTHOR]

In world practice, the role of reproduction of raw material base of oil production by implementing modern methods of oil recovery enhancement (thermal, gas, chemical, microbiological) on the basis of innovative techniques and technologies is rapidly growing and is becoming more important. It is concluded that at present, the priority of increasing oil reserves in world oil production is the development and industrial introduction of modern integrated methods of enhanced oil recovery, which can provide a synergistic effect in the development of new and developed oil fields. This article presents a review and comparative analysis of theoretical and practical methods of improving oil recovery of conventional and unconventional reservoirs. The paper examines in detail methods of improving oil recovery, taking into account the factors of enhanced oil recovery of oil reservoirs. Considered the main methods and technologies currently used to develop oil fields and recommendations for their effective use, taking into account the variety of external factors of oil production: the geological structure of the reservoir, its volume, and properties of oils. It is shown that there is no universal method of oil reservoir development, and it must be chosen after a thorough feasibility study among several proposed models. When describing the methods of enhanced oil recovery, special attention is also paid to the physical processes that occur as a result of applying the technology. In conclusion, the positive and negative characteristics of the presented methods included in EOR are presented, and recommendations that may influence the choice of practical solutions for engineers and oil producers are given. Conclusions are made that development systems, placement and choice of operating mode of wells essentially depend on the geological structure of the reservoir, its volume and properties of oils. An important role in this is the construction of a geological model of the production facility. The used hydrodynamic models of development are based on physical laws, about which oil producers sometimes don't even suspect, and the authors of the models are not always able to convey it to the real producers. The authors consider it reasonable to make a logical generalizing conclusion that understanding processes occurring in the reservoir and taking appropriate measures for optimization and intensification of oil production will allow making oil production as effective as possible. [ABSTRACT FROM AUTHOR]

The article suggests the use of indirect measurements to control the process of induction soldering. Based on the equations of thermodynamics, it is proposed a scheme for approximating the temperature values in the solder zone on the basis of information from pyrometric sensors aimed at an area remote from the soldered joint site. A model-algorithmic instrument for indirect temperature measurements in the soldering technological process is developed, the scheme of which is presented in the article. The software of waveguide paths induction soldering control in the form of an already existing system module has been developed, which allows not only to carry out experimental studies on control algorithms, but also to implement a full-scale experiment, the results of which confirm the effectiveness of the proposed approach.

The aim of this work is to develop an intelligent method for controlling the induction soldering process of spacecraft's thin-walled aluminum waveguide tracks. This new control method is based on the use of intelligent data analysis in order to determine the optimal control algorithms for product heating and the workpiece motion control with optimal coefficients based on the operational data of the induction soldering technological process. The paper presents a mathematical formulation of the intelligent control problem. As part of this work, such intellectual methods have been chosen for comparison as a method based on a fuzzy logic, artificial neural networks, and a neuro-fuzzy controller. Based on experimental data on real technological processes of induction soldering, a comparative analysis of the effectiveness of the proposed approach has been carried out. The results of experimental studies have shown that the control method of the induction soldering technological process based on artificial neural networks has the highest efficiency. In addition, based on experimental data of real technological processes, the most effective structure of an artificial neural network has been determined. The highest recognition accuracy has been provided by an artificial neural network with five artificial neurons on each one of five hidden layers. Experimental quality control of the obtained artificial neural network has been carried out on three different waveguide tracks assemblies with different tube thicknesses: 58×25 mm, 35×15 mm, 19×9.5 mm. The results of the experimental verification have showed that for all the three waveguide tracks assemblies high-quality soldered joints have been obtained. The integrated application of intelligent technologies for controlling the technological process of induction soldering of thin-walled aluminum waveguide tracks could significantly improve the quality of permanently formed connections, as well as eliminate errors associated with human factor. Further research will focus on the intellectualization of other technological processes for creating permanent connections, such as electron beam welding, diffusion welding, etc. The results of this work clearly show the effectiveness of the proposed concept of intellectualization, so it is advisable to extend it to a wider range of technological processes. [ABSTRACT FROM AUTHOR]

This article is aimed at defining and studying the ways to create new methods for determining reliability indicators (RIs) of oil and gas equipment and improving the existing ones. The currently used method for determining RIs in oil and gas industries is analysed. The most appropriate technique for monitoring the reliability of process facilities, adapting to the type of equipment considered, is discussed in detail. Various methods and options for improving such a technique are developed. The application of the obtained methodology results in information on the technical condition of the monitored facility and assistance in determining the timing and degree of technical intervention, which allows enhancing the operational efficiency and, consequently, increasing the service life of oil and gas equipment. The proposed method was approved in practice using a centrifugal pumping unit, resulting in the most effective proposals for improvement, and testing the feasibility of its implementation at oil and gas enterprises. [ABSTRACT FROM AUTHOR]

The purpose of this study is to improve the operational reliability of oil storage tanks and storage equipment at petroleum refineries by extending their mean time to repair (MTTR). The principal methods for studying the problem are the simulation experiment and the statistical processing of the quantitative results of the study. Based on the results of the wall thickness test for the first ring of an oil storage tank, the authors made a case for a regular removal of deposits from the tank walls. The suggested system for tackling the asphaltene deposits (ADs) consists in treating the deposits with a petroleum product heated to their melting point. The optimal solution temperature was chosen based on the results of the tests run to observe the impact of heat on the deposits and establish their melting temperature. A series of experiments were undertaken to identify the optimal composition of heating oil for treating ADs. The methods proposed in the study can be used to build a mathematical model of an AD removal unit for cleaning oil storage tanks, to calculate the operational characteristics of the equipment, and to design a process flow diagram that can be used at fuel-oriented petroleum refineries. [ABSTRACT FROM AUTHOR]